首页 | 本学科首页   官方微博 | 高级检索  
     

基于图形硬件的双调排序算法优化
引用本文:宾洪斌,何锫,胡明辉.基于图形硬件的双调排序算法优化[J].计算机工程与设计,2008,29(14).
作者姓名:宾洪斌  何锫  胡明辉
作者单位:长沙理工大学计算机与通信工程学院,湖南长沙,410076
摘    要:介绍一种新的并行排序算法,该算法以双调归并排序为基础,运用图形硬件的并行体系结构和二叉排序树数据结构的优点,用部分并行代替所有阶段的顺序执行,对双调排序算法进行优化.对该算法进行分析,在理论上n个序列在P个流处理器上的排序,最优的时间复杂度为O((nlogn)/p).实验测试结果表明,优化后的算法比其它基于图形硬件的双调归并排序算法所用时间短.

关 键 词:双调归并排序  流计算  图形处理器  归并  算法优化  排序网

Optimal bitonic sort for graphics hardware
BIN Hong-bin,HE Pei,HU Ming-hui.Optimal bitonic sort for graphics hardware[J].Computer Engineering and Design,2008,29(14).
Authors:BIN Hong-bin  HE Pei  HU Ming-hui
Affiliation:BIN Hong-bin,HE Pei,HU Ming-hui(School of Computer , Communication Engineering,Changsha University of Science , Technology,Changsha 410076,China)
Abstract:A novel parallel sorting is presented.This algorithm is based bitonic merge-sort scheme.The benefit of using GPU's parallel architecture and a binary tree improve the approach.Instead of a completely sequential execution of all stages,the approach is executed them partially overlapped.For sorting n keys utilizing p stream processor units,this approach achieves the optimal time complexity O((nlogn) /p).The approach performs very well in practice,which is caused by several optimizations.Furthermore,the approa...
Keywords:bitonic merge sort  stream computing  graphics processing unit  merge  optimal arithmetic  sorting network  
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号